Local cAMP dynamics in the SERCA2a signalling complex

Abstract

3’,5’-cyclic adenosine monophosphate (cAMP) is a ubiquitous second messenger that regulates multiple physiological functions by acting in distinct subcellular microdomains. Over the last few years, several targeted biosensors have been developed and used in cell lines or neonatal cardiomyocytes to investigate the molecular mechanisms behind cAMP compartmentation. However, it is unclear whether such biosensors can be successfully used for expression in vivo, especially in the context of disease such as cardiac hypertrophy. Importantly, cAMP regulates cardiac function by acting in distinct subcellular microdomains which are independently regulated and confined from the bulk cytosol. Today, this phenomenon is a well accepted paradigm known as cAMP compartmentation. In the heart, one of these microdomains is believed to be located around the sarcoplasmic/endoplasmic reticulum calcium ATPase 2a (SERCA2a). SERCA2a is crucial for diastolic calcium (Ca2+) reuptake and is negatively regulated by phospholamban (PLN). cAMP binding to PKA leads to increased PLN phosphorylation thereby relieving the inhibitory effect of PLN on SERCA2a. Interestingly, SERCA2a expression and activity are known to be downregulated during cardiac disease but cAMP dynamics in such microdomains and their alterations in cardiac disease such as hypertrophy are not well understood. Therefore, the first transgenic mouse model expressing a cardiac specific SERCA2a targeted fluorescence resonance energy transfer (FRET)-based cAMP sensor, namely Epac1-PLN, has been developed in this PhD study. Freshly isolated adult cardiomyocytes of the transgenic mouse line have been used to directly monitor cAMP with high temporal and spatial resolution within the SERCA2a microdomain. To understand the molecular mechanisms that confine the SERCA2a microdomain from the bulk cytosol, FRET results gained in Epac1-PLN cardiomyocytes were compared to those obtained in cardiomyocytes expressing the cytosolic cAMP FRET sensor Epac1-camps. In healthy cells, local cAMP levels in the SERCA2a microdomain after β-adrenergic receptor (β-AR) stimulation were ~4-fold higher compared to the bulk cytosol, which was due to direct phosphodiesterase (PDE)-dependent receptor-microdomain communication. Under basal conditions (in the absence of β-AR stimulation) PDE3 and PDE4 were crucial for confining the SERCA2a microdomain from the cytosol. However, in cardiac hypertrophy induced by transverse aortic constriction, the local basal PDE4-mediated cAMP degradation was significantly diminished, while the cytosolic cAMP dynamics were altered only after β-AR stimulation. Strikingly, local cAMP degradation but not whole-cell changes in PDE activity in hypertrophy led to a dramatic loss of receptor-microdomain communication. In this study, the biocompatibility of the targeted Epac1-PLN biosensor and its potential for real-time monitoring of compartmentalized cAMP signalling in adult cardiomyocytes isolated from healthy mice and from in vivo cardiac disease model was confirmed. In particular, data show that real-time dynamics of cAMP in the SERCA2a microdomain are vastly different from bulk cytosolic cAMP due to local PDE effects and direct receptor-microdomain communication. In cardiac hypertrophy, these processes are dramatically altered which might explain impaired regulation of SERCA2a activity in disease. Ca2+ and cAMP play a critical role for cardiac excitation-contraction-coupling and are known to interact with each other, for example via Ca2+-dependent modulation of PDE1 and adenylyl cyclases 5 and 6 activities. Currently, many FRET studies analyse cAMP signalling and its regulation in resting cardiomyocytes devoid of electrical stimulation to avoid contraction artefacts during the FRET measurements. However, it is not known how such data are comparable with the behaviour of cells under more physiologically relevant conditions during contraction. In this thesis, cAMP-FRET responses to β-AR stimulation and PDE1 inhibition were directly compared in resting vs. electrically stimulated adult mouse ventricular cardiomyocytes expressing Epac1-camps. Interestingly, no significant differences in cAMP dynamics could be detected, suggesting low impact of rapidly changing Ca2+ concentrations on cytosolic cAMP levels associated with β-AR signalling measured with this FRET sensor. On the other hand, after direct adenylyl cyclase activation, PDE1 contribution to total PDE-mediated cAMP hydrolysis increased significantly in field stimulated cardiomyocytes. This could be mimicked by pretreatment of the cells with Ca2+ elevating agents under resting conditions. However, since β-AR stimulation reflects the more physiological situation that is used in the FRET experiments to analyse PDE contributions to cAMP hydrolysis, the use of resting cells for FRET-based cAMP measurements can be justified.